t»yipin >^ oi '-i ' s> • 7f;e PC lineal ^cow^-i ^ // a ^ r(-(9^ i mo ^ Ww^ i ' ^ Rsp C , s ^- ^^ Since the basic premise of this discussion is that ethnicity and ethnic relations are variables which respond through time in accordance to a more basic poTitical-economic infrastructure, it would be misleading of me to approach my topic here as a study of how the Harari urban minority survived in the face of the overwhelming numberica l superiority of the surrounding Oromo. Rather I am concerned with the forces and influences which shaped this relationship and—- perhaps more importantly—which shaped the experiences of these two neighboring populations of the southeastern periphery of the Ethiopian Empire. My earlier analyses of Harari-Oromo interrelationships were reasonably adequate presentations of a system of ethnic stratification which I abstracted «r during fieldwork done from 1962-64 (Waldron. 1974:260-299; 1980:249-256). Their • purpose was to discuss the means by which the Harari townsmen had adhered to thei'" ethnic identity white engaging in daily contacts with other ethnic groups. Beyond that, my basic problem could be described as that of reporting the social morphology of Harari social institutions and their intricate influences in ordering the daily life of the people of the city of Hara'r. At the time of fieldwork, the Harari dominated the landholding in the region immediately surrounding the city. I was able "to summarize some of the patterns of interaction between Harari and Oromo in the context of ethnic stratification. I will list the basics of that abstraction below and then I will * explain some of my , present reservations concerning the adequacy of the earlier approach to Harari-Oromo relationships. ^^^ 1. The Harari occupied the position of a minority elite class in local system of stratification vis-a - vis the surrounding Oromo =^^= ° majority. social system is said to be stratified if one class dominates the ^^^ S3 other(s) by virtue of the dominant group's control over some essential resource (^hjJj}) the subordinate group needs to survive (cf. Fried, 1957:186). The Haran-Oromo system existing in and around the city of Harar was dominated by the Harari because they owned much of the agricultural land in this area. The Oromo of the region had access to farmland, essential to their economy, only under systems of tenant-farming or sharecropping (Yusuf Ahmed, 1961:23) which were '. very profitable to the Harari, but whose rent fund was exorbitant for the Oromo fanners. 2. This system is appropriately deemed ethnic stratification (Cohen, r974:xviii) inasmuch as ethnic differences coincided with class differences. That is, in this concept, the Harari ethnic group comprised the dominant class while the Oromo peasantry constituted the subordinate class. Membership in the upper class thus was manifested by ethnic identity, which in turn separated Harari from Oromo. Specificall y, the Harari spoke their own distinct Semitic language (ge sinan), practiced their own particularly complex social customs (ge^l ZBJI . and^ physically separated themselves from the Oromo by ng inside the wall of their city of Harar which they cal ( Ti ^^^ T'ge. T ^ hey called themselves ge 'usu. "people of the city." ——3. The Harari dominance of this system can also be viewed as a set of Harari-managed transact!onal systems. Membership within the dominant group was monopolized by a strongly stated ideal of group endogamy which was statistically upheld in practice. Information between Harari and Oromo flowed freely, but can be viewed as being under Harari control, since Harari were bilingual in the Oromo language, but Oromo rarely spoke Harari. Goods and services were exchanged between the two ethnic classes although almost always under conditions which favored the Harari. The economy of the Harari depended upon Oromo labor, particularly under the system mentioned above. Harari also profited from Oromo market transactions, since Harari were the "local merchants who controlled the market and established prices. Thus this system persisted because the Harari permitted the exchange of goods, services, and information across the ethnic-class interface, but virtually prohibited the Oromo from acquiring Harari ethnic identity (cf. Waldron, 1974:276). Notice, in this system of ethnic stratification, that the Harari, by preserving their ethnic identity, also virtually blocked Oromo advance in social status to the dominant class. This brief summary gives the outline of the early model of ethnic stratification, although the entirety is more complex. In these earlier presentations, the uniqueness of the Harari one-city culture has tended to obscure its comparability. Harar can be viewed as a variant of the Muslim preindustrial city (Sjoberg, 1960:52 ff.; Waldron, 1979:239-58), and in this approach, the Oromo can be viewed as the peasant population whose labor support!the specialized mercantile, administrative, and religious services of the Great Tradition of urban Muslim Harar. now feel that the foregoing model of Harari-Oromo ethnic stratification, useful as it is in the limited sense of synthesizing data observed in the 60's, has several significant shortcomings. 1) By failing to explain the role of outside powers, particulary Ethiopian rule, it greatly oversimpli fied a complex reality. 2) By failing to account for the political and economic factors which, as I will explain, evolved through time, it reifies, i.e., makes overly rigid and concrete, the relationships of the two groups. 3) Finally, like many synchronic studies, it seems naTve in the light of historical evidence. can conclude this introduction by making an assertion which seems contrary to the model of ethnic stratification, and which the main body of the discussion will (clarity./ The Harari did not, in any simple sense, retain their identity and other aspects of their social system through the centuries of contact with the Oromo. Rather, the Harari and the Oromo of the region differentiated and developed their social characteristics—including ethnic identities'-in the context of a political-economic system which became more and more unitary as time approached 1875, but which underwent divergent differentiation under Turco-Egyptian rule (1875-1885) and, more markedly, under the domain of the Ethiopian Empire (1887-1974). In this perspective, the early model of ethnic stratification appears to be but one facet reflecting the underlying pressures of one point in time, rather than a definitive statement of Harari-Oromo interrelationships. I. 1559-1875: Three Centuries of Negotiated All iances Initial Contacts The context for the formative phase of Rarari-Oromo relationships was established by the two major events in sixteenth-century Ethiopian history: the jihads (1529 - 43) of Imam Ahmed ibn Ibrahim al-Ghazi . called Gragn ("the left-handed") in Amharic tradition, and the subsequent demographic explosion of the.Oromo from their homelands in southern Ethiopia. The social, economic, and political disruption that the jihads and their countermeasures caused is well - documented (Shihab ad-Din, 1897; Trimingham, 1965:91-93). The first reported appearance of the Oromo in the region of Harar, which was then capital of the 'Ada! Empire, was an encounter between a gada force and Amir Nur, Imam Ahmed's successor in 1559 (Cerutii, 1931:57). The consequence of this; initial contact was a devastation which established a major component of these and future relations in Harari perceptions: fear of Oromo aggression. Until this time, the arable regions of the Harari highlands, extending in the east to the present range of the Barteri and Geri Soma1icu1 tivatiors ^) and < in the west well into the Ch'erch'er range of the rift projection which connects Harar with the Ethiopian Highlands, seem to have been occupied by a proto-Harch farming population (Cerull i, 1936:441; Braukamper, 1981:86) who, tike modern Harari, had dose linguistic affinities with the Seiti Gurage (Hetzron, 1972:42) and who may have been an ethnic offshoot of that group ' s predecessors. Pending archaeological research, the full extent and economic base of the proto-Harari population is unknown. Their last trace, however, is the record of their decimation. * ...by 1577 (the Oromo) had destroyed more than a hundred villages and beseiged the city, "Until the gates were fill ed with corpses". (Muhammed Hassan, 1973:iv, citing Ms. ^300, Institute of Ethiopian Studies; parentheses added). Amir Nur had reacted to the Oromo invasion by building the city ' s wall in 1567, providing a refuge for surviving farmers and the established residents, thus establishing Harar as a one-city culture, an immured and encapsulated society whose survival, in these early days, was extremely precarious. The - Sultanate of 'Adal, it should be noted, abandoned Harar and moved its capital to the Danakil oasis of Aussa in 1577. There was no question of Oromo military superiority. Oromo manpower was organized, not in terms of a conventional military hierarchy, but in terms of the very social orgiimzation which defined their indentity. --Behind this great spread of (Oromo) society there lies an institution--the Gada System: which was at once the basis of t. heir military conquensts and which set limits on demographic antecendants.... The Gada system is a system of classes (luba) that succeed each other every eight years in assuming military, economic, political and ritual responsibilities.... Before assuming a position of leadership the (dominant) gada class is requf|JJed to wage war against a community • that none of their ancestors had raided. This particular war is known as butta and is waged on schedule every eight years. It is this . event that was most directly connected with the pulsating frontier of their dominions in the sixteenth century leading toward the conquest of nearly half of Ethiopia's land surface (Legesse, 1973:8). The Harari were small in numbers compared to the Oromo, probably reduced to around ten thousand, having suffered from war and a three-year famine (Trimingham, 1965:93). They were restricted to the city, deprived of the.'' former Somali allies of the 'Adal Sultans, and almost certainly .cut off from the lines of supply to the Red Sea coast which had provided a small number of firearms and mercenaries trained in their use in earlier days. By virtue of its consolidated position behind the city wall , Harar was able to withstand attacks of localized lineages of Oromo, but would have coll apsed in the face of a concerted gadabased large-scale attack. » Unlike Fatagar, Haddiya, Dawaro and other principalities in the path of the Oromo expansion, Harar survived, probably because of its market convenience for the Oromo (Muhammed Hassan, 1973:1). The first recorded negotiation between Oromo and Harari was a short - lived market pact in 1569. which guaranteed the Oromo . , cloth at a fixed price in the markets of Harar. Cloth had been a mainstay of the trade of the Muslim principalities for centuries but, for the Oromo of the period, It was probably a sumptuary good, a replacement for hide wraps. In my opinion more important commodity, whose local monopoly by Harar's merchants may well have tipped the balance in favor of the city's survival, was salt, whose sodium con-tent provides an essential human nutrient. Bloch has established the critical role that monopolies in salt trade have played in political centralization in the ancient Near East (Bloch, 1963:88-96). Rathje has suggested that control of salt trade was one of the primary economic bases of the rule of the Yucatecan Mayan theocratic elite (RathJe, 1971). Abir, noting the absence of salt or other suitable sodium sources in the Ethiopian highlands, similarly cites its monopoly by the rulers of 'Ifat and other early Muslim trading centers as crucial to their power (Abir, 1980:8). For the Oromo of the region, Harar was the provender of salt, and the Harari had the specialized mercantile knowledge and organization to bring it from the coastal sources. The task of the Harari amirs during this initial period of contact was formidable. They had to find means of opening trade routes to the sea, through Oromo-held territory, and, beyond extending market privileges, they had to discover means of establishing alliances with local Oromo in order to prevent a mass attack on the city. Caravan trade through Somali territory was managed by hiring an abban from each lineage, who acted as a trade-protector through his own territory (Yusuf Ahmed, 1961:37). The Oromo, however, had no such experience with trade, nor had they any established social role suitable for trading pacts. Discerning an institutional means for establishing trading and other all iances with the Oromo presented the amirs of Harar with a serious problem upon which the city's survival hinged. The problems of the Harari amirs in their search for alliance partners was complicated by structural differences between the two societies in contact. Harar, small as it was. was organized in bureaucratic manner. Under the amir were categories of officials—whose specific titles and duties will be summarized below—who connected him with the citizenry, primarily through taxation. Contrac^ i/ obligations were understood and recorded in Muslim shari'a courts (Stitz» 1975:9). Oromo society, however, was essentiall y kin-based and egalitarian. 'Local organization was based on the patrilineage (gosa), a characteristic principle of social organization of many East African pastoralists. Unique among the Oromo, however, was the gada system mentioned earlier, which culminated the local organization of the gosa in an elected council whose members served eight year terms (rather than being "chiefs" or other permanent representatives of the polity). The Oromo who came to Harar comprised two gada-based confederacies, the dominant Afran Qall o, made up of the Alia, Nole, Obora, and Babile and the Anniya. As Muhammed Hassan has reported (1973:23-25), the gada-like ruling council of the Afran Qallo was called the Roba-Dori. The highest titles therein were the Abba Fugug, or president, and the Abba Dula,or war leader, who could call up the population-wide military force, the Dori. As stated, the personal powers of such positions' occupants were limited by the eight- year tenure, and they were further regulated by an pral constitution, the Hera Gosa. The constituent patrilineages were represented by Bokku,an office which Legesse translates as "ritual leader" (1973:69) and which Muhammed Hassan, (1973:24) referring specifically to the Afran Qallo, calls "spokesman". Despite the specialized duties of the officials described here, they were limited, both by structure and by Oromo ideology, to a status of primus inter pares. The Oromo system, as initially experienced by the Harari, provided no structural equivalents of the amirs, who were the elite of a stratified society. ' The quandary for the amirs, then, was to find alliance partners who could represent the body politic of the Oromo, in the context of an institution which would substitute for a contracted relationship. Since the Harari were on the defensive, they had to respond to Oromo institutions, rather than vice versa. Thus it seems as though the only institutions available to the amirs for pact formation were those defined in terms of kinship: marriage and adoptive brotherhood. Interethm'c All iances and Interdependency AHarari document referring to the reign of Amir Abdull ahi Ibn Ali (1671-1700) testifies that marriage between the sons of amirs and the daughters of politically influential su/rounding groups was an established practice in this early period of contact (Cerull i, 1942:5). Such strategic marriages by the amir garach ("amir's households"), which contrasted strongly with the ethnic endogamy of Harari commoners, prevailed throughout the history of the Harari amirate. The structural implications of these marriages were defined by the reciprocity entailed in Oromo affinal ties. In 'exchange for his daughter, an Oromo all iance partner could expect periodic gifts from his in-laws. In the context of pastoral Oromo society such affinal reciprocity would serve as a link between gosa which, by circulating wealth, would provide a levelling mechanism within the society as a whole. As applied to the amirs of Harar, the relationship probably assured them of defined and peaceful relationships with a specific gosa^paid for in terms of market wealth and taxes collected by the amirs from Harari citizenry. In this manner, the early amirs were able to negotiate a degree of peace and safe passage with a few Oromo groups, at least for limited periods. The drain on the city' s resources would be commensurate with the number of such marriages. The number of unaligned—and thus potentiall y hostile Oromo groups--wou1d probably alwavf outnumber those who were so allied. These a/liances, as they developed through time, had important ramifications for the political economy of the region. By the mid-nineteenth century, Oromo all iances had become intrinsic in succession disputes of the Harari amirate (Caulk, 1977:378-9), wherein an amir would have a certain group of Oromo all ies, but.his contesting claimants would, through their own strategic marriages. develop mutually-exclusive pacts. In these later years of the ami rate» Oromo might fight Oromo because of such conflicting alli ances. The economy of the city vacillated according to which of its trade routes were open at a given time, and the security of the city's provisioning gardens, adjacent to the walls, would vary as these alliances waxed and waned. Since the amirs were effectively purchasing the city's freedom under conditions of negative reciprocity (Sahlins, 1972:272), the Harari citizenry progressively came to resent the amirs and their taxes, as well as the Oromo who ringed the city. The former resentment, believe, conditioned the Harari attitude toward government in general. The latter negatively conditioned Harari stereotypes of Oromo. During the same period, local Oromo society was also transformed by the influx of Harari wealth to the recipient marriage partners. A sufficient motivation for Oromo patrilineal families' participation in transgenerationany continuous intermarriage was provided by the promise of gaining continued access to the new forms of wealth and privileges (Burton, 1965:174) thus obtained. Permanent differences in wealth developed within the local Oromo population, realized in terms of luxury goods obtained from Harar's markets, expanding herds (Caulk 1977:374) demanding herdsmen's labor far beyond traditional gosa cooperative potentials, and land grans from the amirs whose potential could be best realized through the adoption of Harari agriculture. These new forms of wealth produced an emerging elite among the formerly strongly egalitarian Oromo. Muhammed Hassan states that "By the 1870's. most of the Afran Qall o and Ettu Oromos were sedentary agriculturalists il wh e most of the Anm'ya were pastoralists (1981:2). (Among the Anniya were such men as Mayu whose wealth in cattle is still proverbial among the Oromo (Muhammed Hassan, 1972:10). Ideally, wealth was measured "in terms of cattle owned. Owing to the transformation of the Ittu and the Afran QaTIo from pastoralism to sedentary agriculture, land also became a measure of wealth. As a result there were rising "landowning classes known as Abba Burka ("father of springs") among the Ittu and Abba Lafa ("father of land") among the Afran Qallo. This phenomenon undermined the traditional political organization from within. Instead of able leaders elected for their qualities, the higher offices were filled by landowning nobilities (Muhammed Hassa^ 1981:4). This emerging Oromo elite provided, finally, a perrnaneTrt and identifiable set of potential alliance partners for the amirs. Their progressive involvement in the market economy shifted their orientation toward the Harari sphere of influence and away from the traditional Oromo economy and obligations. At the same time it gave them a new power over their subordinate kinsmen. The amirs of Harar consolidated this orientation by extending land grants to the nascent elite and by awarding them tax-collecting privileges and titles under the Harari administrative system. The Harari administrative system, primarily concerned with taxation, was simple hierarchy , within the city, each of the " five quarters was under the charge of a malaq, who had, in turn, several lesser officials called garad under him. As the system was extended (probably no more than a twenty-kilometer radius) outside the city, the hierarchy proliferated. Referring to the early nineteenth century, Yusuf Ahmed says, * garad is the chief of a vill age, or sub*-vi11 age; the damin is the chief of a whole tribe. Several garadach*, sometimes five or six, come under one damin. The garad or the damin, each on his own "level, is the Amir's administrative agent; he distributes justice, and coll ects taxes or tithes. At the head of the above mentioned chiefs is the dogin (Yusuf Ahmed, 1961:23). Although, under the amirs, the highest officers were always Harari, many officers below the level of dogin were the aforementioned Oromo elite. Besides receiving -. an estate in land from the amir and regional monopolies in taxation, "the bearer of the office enjoyed the free service rendered to him by the peasants under his garadship or daminship" (Yusuf Ahmed, 1961:24). In-this manner the amirs CO-Qpted the Oromo elite. Oromo and Harari Understrata The-emergence of the'Oromo elite and their gradual adoption of agriculture was accompanied by a further differentiation of Oromo society: the appearance of an Oromo agricultural peasantry, referred to categorically as Qottu ("till ers of the soil"). Partiall y, the formation of this peasantry can be understood as a reflex effect of elite formation. New demands for labor, new forms of marketeer!ented wealth, and new bases of power among the elite all contributed to the processes of sedentization and peasantization of those in a subordinate position. Closer tp the city, in the Harari farms and orchards, Qottu also provided most of the labor by the nineteenth century. I believe that demographic pressures exacerbated by local climatic variations probably explain the gradual development of qottu dependency upon Harari landlords more reasonably than any explanation which suggests that Oromo labor was controll ed by Harari force. In the 1960's I observed numbers of ' vp^ooked .'. Oromo—essentially refugees from local crop failures—congregating in the Faraz Magalla as beggars. At that time, Harari were innovating a practice of daily *hiring, and utilized these Oromo as a labor pool. In earlier days, and to some extent in the 1960 ' s ^ Harari would select some of these Oromo as haras hi ("farmers"), who would be established as laborers and watchmen on Harari farms. They received food and shelter, but no wages or share of the crop. They were free to return to their vill age of origin, but many served for years. Yusuf Ahmed also cites patterns of tenant-farming (oyna) and .sharecropping (ha rta) used by Harari in regions farther from the city (Yusuf Ahmed, 1961:25). The Harari citizenry—those who were neither members of the amir garach nor members of the ami^s government—also formed a kind of understratum during the period which ended in 1875. Unfortunately, social and economic data concerning the general citizenry of Harar is virtually absent for this period. However we know three important facts concerning them. First they were socially distinct from the amir's family, a distinction which is still acknowledged (Maldron, 1974:305 - 313), although its ^ importance lapsed with the end of the amirate. Second, they paid the bulk of the taxes which supported the amir's government; including his alliances. Third, they become progressively resentful of the later amirs and the privileges extended to the Oromo elite and their kinsmen. This resentment culminated under the reign of Amir Muhammed (1856-1875), who is reviled in the oral tradition of the city for having entered the i1 man gosa (adoptive brotherhood) with the Bokku of the Alia Oromo. In order to finance the extensive demands of his adoptive kinsmen, Amir Muhammed coll ected the established zikat,a ten-percent tithe, but also devalued the city's currency while estab-'' tishing a special mahalaq al-Galla or Oromo tax (Yusuf Ahmed, 1961:35). Stitz has documented the poverty Qf the ordinary Harari citizens during the nineteenth century (1975:1-11). As a direct result of the resentment and alienation evoked by Amir Muhammed,- a contingent of Harari invited the Ottoman Turco-Egyptians to take control of Harar. which.they did in 1875, thus ending the independent amirate. Ethnic Relations and Stratification: 1874 At this point, I would like to return to - the original model of ethnic ^ stratification and compare it with a summary of political organization and ethnic relations between the Harari -and Oromo at the end of the independent amirate, on the eve of Turco-Egyptian occupation in 1875. The Harari may be conceptualized.as Internall y stratified, with the amir • garach separated from the populace at large. Aside from the close control of succession by the dynasty of Da'ud ibn 'Alt, within the amir garach, this separation was more a matter of activated status claims than being a matter rigourously required by Harari society. In fact, data obtained during the periods of fieldwork are somewhat ambiguous about whether the amir garach, as sizeable network of households was a self-contained, endogamous entity, or If they intergraded through marriage links with the rest of Harari society. believe that the households most closely related to the ruling amir would be those occupying high status and who would maintain the greatest social distance 3 - X from the general citizenry (Maldron, 1974: ^ ). The links between the two ethnic groups . are traced through the elite of each: political ties were kinship ties, either marriage or adoptive brotherhood, between the amirs and the emerging Oromo elite. The Amirs of Harar would provide lucrative prestations, founded on such ties, by taxing the Harari populace. Thus the city ' s vital prerequisites, peace and open trade routes ^ would be measured by the success and extent of such ties. These would be individually negotiated by each amir, although transgenerationat continuities would be expected, particularly where such alliances controlled vital trade routes or militarily vulnerable points. To the 'extent that these alliances were successful, the citizens could expect to be taxed and otherwise economically burdened. That is, the more peace, the more poverty, unless trade volume compensated for the payoffs to Oronio allicince partners. ^ Representation of Harari interests could be accomplished in a unitary fashion by virtue of the single office of amir as spokesman for his people. Oromo representation probably always was particularistic, inasmuch as even elite Oromo retained gosa obligations. Despite the attempt to encompass the Oromo elite in their administrative bureacracy, Harari political relations were always manifold and multiplex. Only Amir Muhammed attempted to unify this representation by entering the adoptive brotherhood which effectively connected him, en bloc, with the entire Afran Qallo (Muhammed Hassan, 1973:13). If this is considered as a daring diplomatic maneuver (Caulk,/ f7 / : ^7f }, rather than as a sell-out to the Oromo, as Harari traditions insist, the dilemma of Amir Muhammed can be seen clearly: He consolidated his alliances with the Afran Qatio via the ilman gosa brotherhood, which was politically sound. However, the price (Muhammed Hassan, 1973:15) was a vast proportion of his personal income and the income of the citizens, which was economically unsound. The more he succeeded the more he failed. Unifying the two ethnic groups would have eased his dilemma, and he tried to do this, again through the only institutional option available, Islam. Amir Muhammed tried to extend the unifying principles of Islam throughout Oromo territory (Muhammed Hassan, 1973:18) but met with only limited success. Instead of unifying the Harari and Oromo, he deeply alienated the Harari, who perceived him as an Oromo, and whose perception of Oromo as aliens and enemies was,probably heightened by Amir Muhammed ' s efforts - The Harari general citizenry was displaced from the all iances which were vital to the city's survival. Thei-r lives were more directly affected by the increased economic burdens induced by effective.alliances, and the conditions of escalating levels of warfare with Oromc- when all iances were not effective. The amirs and the Oromo were to blame for both in the eyes of the Harari citizen. The Harari citizenry of pre-occupation nineteenth century dealt with Oromo •i peasantry, primarily sedentarized Qottu, in two contexts: farm labor and other service occupations, and in the market as customers. Both of these contexts involved the Harari in dominant positions and Oromo in subordinate ones. Moreover, Harari viewed their own Muslim urban society as superior to that of the Qottu, and this accentuated the stereotype held by Harari of Qottu as lowstatus, potentially dangerous people. Exacerbated by the Harari blaming of Orcyno for most of their troubles, a pattern emerges which provides a degree of understanding for the attitudes upon which intergroup relations rested in later decades. Like so many intergroup prejudices, Harari stereotypes of Oromo were irrational simplifications which overlooked the underlying political and economic determinants of the social predicaments of both groups, and reacted by blaming the tower status group for the perceived troubles of the upper status, dominant group. Qottu perceptions of Harari as hard-hearted exploiters was the reciprocal of this stereotype. In summarizing the social developments of the first three centuries of Oromo - Harari contact, one must first acknowledge the virtual absence of social data. The notable exception is Muhammed Hassan's work, upon which I have heavily relied. From the available evidence, it seems clear that the contact of the two groups had reached a point, by 1875, of incipient stratification. believe the stratification process was incipient, rather than complete for two reasons. First, ethnic identities and obligations, to a degree, continued to cross-cut, and to compete with, the mutual interests which were otherwise uniting the Harari-Oromo alliance partners. Second, and augmented by the previous statement, despite systematic attempts .to enlist the Oromo elite in their administrative system, the Harari amirs never succeeded in consolidating secure power base. There were really, then, two ethnic elites with some interests in common, but many points of contention. Although commercial, administrative, and religious institutions centered on the city of Harar, power in this evolving system ultimately rested in the hands of the Oromo, thus circumscribing the freedom of movement of the Harari populace, and Involving the amirs in a never-ending series of tentative negotiations. Stitz (1975:10) suggests that -the Harari-controlled gardens never extended more than five kilometers from the city during the nineteenth century. Muhammed Hassan (1973: 28) suggests that the hegemony of the Harari administrative system never extended for more than twenty kilometers radius from the city previous to the occupation. An Oromo friend summarized the locus of power in this system, saying, "My parents said that in the old days, if we (Oromo)- did not have our demands satisfied, we would surround the city so that we could be seen, until the Harari gave in." One might speculate that, had Amir Muhammed succeeded in his machinations to form a unitary alliance with the Oromo, then the region of Harar might have become an Oromo-dominated society with a unified elite, a truly stratified society which might have provided a major power in the Horn of Africa. However. ethnic identity overrode class formation, inasmuch as the alienation of the Harari citizenry led directly to the invitation extended to the Turco-Egyptians to take over control. A short time after the occupation. Amir Muhammed was strangled {Ualdifen,!9^b:^7/i.)^ bringing to an end this phase of development of Harari-Oromo relations. 11. Ottoman Rule. 1R75-1885: Consolidation of Stratification Despite the brief period of Turco-Egyptian rule, several significant changes in the political economy of the region took place which, in retrospect, may be viewed as pre-adapting the region for the types of occupation later experienced under Ethiopian rule. As will be shown, many of these changes favored the Harari, while the Oromo were supressed. From this time onward, Harar and its environs ceased to be an independently evolving system, and were relegated to the status of a secondary periphery whose fate was directed by a succession of governments and political influences over the next century. Oromo Subjugation Although the arrival of Rauf Pasha and his troops "at one o'clock in the afternoon of October 11, 1875 (Paulitschke, 1888:229) was met with peaceful receptions by. Amir Muhammed and by the Harari who opposed him, the Oromo, correctly anticipating the result of a new source of power in the region, vigorously and quickly opposed the newcomers. First the Nole arose and were defeated (Muhammed Hassan, 1973:22), and, during the next year, several other battles took place, including the last mustering of the Dori. the collective army of the Afran Oatio (ibid.). The Turco-Egyptians fortified the city's wall and built battlements (which have since disappeared) in several places, but 1,he outcome was uncertain until reinforcements arrived from the Reo Sea (Paulitschke, 1888:230), Mohamed Mokhtar Bey, who commanded one of the reinforcing batallions, acknowledged the persistance and bravery of the Don' (Muhammed Hassan, 1973' 2 .^ ). However, for the first . time in the military history of the region, firearms—in the hands of the occupiers—dominated vastly superior numbers of attackers armed with hand weapons. .- Hereafter, in the Horn of Africa, access to • foreign manufactured firearms and munitions became the critical factor in determining the survival of inland political futures. By 1877» Oromo military might in the region was destroyed. The Turco-Egyptians followed up the military defeats of the Oromo with three other types of action which subjugated the Oromo and centered their administration in the city of Harar. The first of these was the elimination of several Oromo leaders, a victorious event in Harari oral history, but which Muhammed Hassan reports was a massacre accomplished with duplicity (Muhammed Hassan, 1973:22). The second action, designed to implant uniform sanctions of control in the region, was a concerted campaign of conversion to Islam, accomplished, where necessary, by forceful circumcision of Oromo men (Muhammed Hassan, 1973:27-28), including those referred to earlier as the emergent Oromo elite. The third process, through which the Turco-Egyptians attempted to routinize their rule of the region, was the extension of a form of the Harari administration and taxation system throughout the region. By the end of Egyptian occupation, the administrative system had been fc / i/CtIft-ti i extended to a radius of 40 from the city, double its maximum under the amirs. The taxation system thus incorporated covered some 200,000 persons, and the.titles of the administration of the old amirate had proliferated (Muhammed Hassan. 1981:5). * ... the Egyptians grouped the people under their administration into villages and appointed 11,829 nereditary Malaq and Garads. Over these they appointed about 500 Dam-ins. At the top of this hierarchy they appointed 45 Dogin who were always Harari (ibid.). In several ways the Harari benefited from the new order, just as the Oromo were reduced in status. The enforced peace permitted Harari to move freely in the adjacent farmland. The overall economy of the .city prospered. Under the amirate, in its last years, only seventy caravans per year brought goods from the coast, whereas under the Turco-Egyptian rule, the city now handled four hundred per year (Pautitschke, 1888:238). With improved sanitation and medical services, as well as the lure of an improved economy, the population of the city Qrew to some 45,000, of whom over 8,000 were Ottoman troops (ibid.). Taxation continued under the new system, but the currency of the city was changed from Amir Muhammed's debased muhaTlaq to Egyptian coins. To house and provision their troops, the Turco-Egyptians carried out the first wave of land alienation, process which was to beset the Harari throughout the next century. However, the Harari gained land during this period, rather than lost. new system of land tenure was introduced by the occupiers, whereby the land was claimed as sovereign property, but could be "repurchased" from the occupiers (Muhammed Hassan, 1981:5). The Harari dogin, damin, and garads were in a position to benefit from this and other innovated policies concerning land tenure. Not only did they have strategic access to the new government, but, since they derived an income from their positions, they were in an optimal position to select and validate land claims. The systematic surveying of farm and urban properties consolidated the position of the Harari versus the surrounding Oromo. Although some Harari ha^ deeds (hojja) which back to the reign of Amir Abd as-Shakur (1783-95), Uli& these were individual records whose preservation hapchance. The intervening amirs did not seem to have carried out comprehensive registrations, although shari'a court records, by proving a continuity of inheritance cases, provided testimony to individual ownership in many cases (Stitz, 1975). These older land registrations certainly did not include the expanded domain from which many Harari benefited during the years 1875-85. It was during this period that Harari expanded their land holdings, registered them in the land office of the city, and kept them active by paying taxes. lHarari capital formation expanded during this period, as a direct consequence. Whereas Stitz ' s preliminary examination of the inheritance records of the Islamic courts from the early 19 th century showed that the average Harari was poor, holding farmlands only up to five kilometers'radius from the city (Stitz, 1975: 9-11), Harari holdings at the end of the 19 th century were certainly multiples of this. Although the social identity of the established Harari long-distance traders as they existed under the ami rate is vaguely documented, one can state with confidence, as well as remain in accord with the memories of elderly Harari, that Harari entered retail merchandising on a large scale only after the coll apse of the ami rate, a process which gained momentum after Meneliks conquest. and is considered below. Some of the Oromo elite who were coopted by the Turco-Egyptian administration benefited from the Turco-Egyptian rule, as far as increased land-holdings were concerned (Muhammed Hassan, 1972:29). However, they did not have the other advantages of the Harari , whose citizenry's endogamy and other boundary -maintenance practices (Waldron, 1974:270-278) tended to exclude them from the city's social, economic and informational systems. Besides the enrichment of Harari rank-and-file, several other changes had taken place under the Turco-Egyptians which altered the relationship of the Oromo and Harari by the end of this decade of occupation. Under the ami rate, the crucial link between the two groups had bee'n the alliances between the amir garach and the Oromo elite. This relationship was supplanted by the establishment of centralized rule in the region, and, of course, the amirate was eliminated, except for the short interim of Amir Abdullahi. The Afran Qall o was broken (Muhammed Hassan, 1981:3), while the enforced peasantization and conversion to ,' Islam of the majority of Oromo in the region expanded rapidly. II . The Short Reign of Amir Abdullahi (1885-1887) With the Mahdist rebell ion in the Sudan, the economic coll apse of the Cairobased Ottoman empire, and the supplanting of that rule by British colonial power, the situation in Harar changed abruptly. For reasons beyond the scope of this discussion, the British decided not to retain Harar. By 1885, before the final withdrawal of Turco - Egyptian troops, the British attempted to prop-up rule in the region, preparatory to their own pun-out, by reinstituting two decadent institutions: the amirate of Harar and the Afran Qatio (after a pledge of peace and harmony with the city (Muhammed Hassan, 1981: , 5 ' )^} Bab Haji Mume Bashir, who was 97 in 1963, remembered the reign of Amir Abdullahi ibn Muhammed (1875 - 1887), saying, " Mir'aj nara" (It was a dream"). The last Turco-Egyptian troops left Harar on May 20, 1886 (Paulitschke, 1888:236), and by January, 1887 Harar had been incorporated in the Ethiopian Empire. Menelik's conquest of Harar was an intrinsic part of his plan of expansion. He had mentioned his intentions in a letter to King Umberto of Italy as early as 1884 (Marcus, 1975:90). He required a reliable southern trade route to the Red Sea ports in order to exchange the booty extracted from his previous conquests of Sidamo and Arussi for imports, particularly firearms and munitions, to fortify his position vis-a- vis Emperor Yohannis, whose power he was contesting, and who controlled the northern routes from Shoa (Marcus, 1975:89). When Amir Abdullahi's army massacred the Italian geographical expedition of Porro, in 1886, Menelik was provided with the rationale necessary to invade Harar white appearing to be a pacifying influence in European evaluations. The particulars of Amir Abdullahi's rule and the battle of Ch'elenqo have been given thorough historical analysis by Marcus (1975), Caulk (1971), and Muhammed Hassan (1981). My comments here will thus be restricted to Amir Abdullahi's symbolism of response to the crises of his time, since it epitomizes - t C.-- the major pattern of adaptation of Harari traditionalists throughout the period of rule of the Ethiopian Empire. By translating urgent political, economic, and social pressures of change from secular matters into a religious symbolism based on Islamic fundamentalism. Amir AbduTlahi was able to deal with the traumatic events of his reign as threats to the purity of Islam in Harar, an effective means of mobilizing the orthodox Harari, but an ineffective means of warding of Menelik's thousands of well-armed troops. By the end of the Turco-Egyptian occupation, Harari society had become socially and politically encapsulated. Under the umbrell a of power provided by the welcome—but foreign—Turco-Egyptians, the city's economic base had expanded and become secure and the chronic threat of Oromo agression was quelled. The city wall thereafter became more of a symbol of social separation from the Oromo and the world at large than a defensive structure (Waldron, 1980: ^° ). Having lost its political independence (except for the eighteen-month reign of Amir Abdullahi), and having been progressively alienated from the machinations of its ami rate during the centuries of alliances with- the Oromo, the citizenry of Harar was displaced from the arena ^political decisions which affected its fate and was distrustful of those who acted on its behalf. As has been discussed elsewhere for the transition period of 1974-75 (Koehn & Waldron, 1979: ^-6^), social and economic pressures of change were perceived by the leaders of the important communal groups of the citizenry, the'afochas, as threats to the Muslim purity of demeanor in the celebration of weddings and other rites of passage. I believe that Amir Abdultahi, who is remembered in Harari tradition as a sofi, a religious scholar who was removed from mundane events, foll owed this pattern and helped establish it as the modus vivendi of the city's inneraFcc^J directed^leaders. Effective as it was in revitalizing Harari culture within the city wall, it was il l-adapted to responding to changes in the political economy of the region. i » " Within this orientation. Amir Abdull ahi devoted his short reign to the furtherance of Islam and its principles, as he perceived them. He spent considerable effort in proselytizing Islam among the Oromo whom the Turco-Egyptians had forcibly converted. His best remembered reforms among the Harari populace were to require that women wear Turkish-style pantaloons (ganafi) and to forbid the sale and consumption of alcoholic beverages, which Burton (1966:/^ ) had cited as being profligate in the pity. That the Harari adopted these reforms probably had more to do with their function as ethnic markers which helped the Harari maintain a discrete identity vis-a-vis contingent outside groups than it had to do with the lasting impact of Amir Abdullahi's fundamentalism. He attempted to expunge Harar of the corrupting influence of non-Muslim foreigners by closing most of the European trading houses which had flourished under Turco-Egyptian rule. , and it was in this spirit that the expedition of Porro met its fate. think it is extremely significant to the theme of adaptation that I am suggesting here, that a Harari document reproduced by Muhammed Hassan (1973:52) interprets the ensuing conquest of Harar as a manifestation of the wrath of All ah over the unjust killing of Perro's expedition, rather than perceiving it as the poorly organized, inadequately manned and equipped military effort that it was, at least in worldly terms. • In the same vein,,Amir Abdullahi's last act as Amir was stylisticall y appropriate. He reacted to Menelik's armies by calling the last Harari jihad. According to Ragazzi, an Italian doctor accompanying Menelik's troops, the Harari jihad coll apsed after fifteen minutes (Ca</U, 1971: 15) and the main resistance to the Ethiopian forces came from the Oromo--Muslim and non-Muslim—who rationally feared for their herds and lands, as well as irrationally hoped for fortunes in booty. Harar was occupied on January 9, 1887 (Caulk, 1971:17), and Amir Abdullahi fled, although he survived as a sofi, seldom leaving the confines of his residence, until 1930. III . Harar, Ethiopia The establishment.of Imperial rule under Ras Makonnen radically changed the conditions of life in the region which was to become Hararghe Province, Ethiopia. new political culture was forcibly implanted. A new population of soldiers, governmental officials, and highland immigrants numbering in the thousands challenged the productive carrying capacity of the region, and new patterns of land tenure, tribute and taxation were introduced to the region. The overall effect-was to benefit the newcomers. Harari landholdings were circumscribed and the Qottu peasantry were virtuall y disenfranchised. Power in the politicaleconomic system was now monpolized by highlanders, indiscriminately call ed " Amhara" in the region, who controlled the government and who progressively expropriated land upon which the sustenance of the pre-conquest populations depended. The agricultural and trade wealth of the" region was largely tunneled into Menelik's treasury and the estates of his officials. The sanctions of the imposed government rested upon its identification with Orthodox Ethiopian Christianity and the myth of Solomonic descent—neither of which were shared by the Harari and Oromo. But behind these "lay the occupiers' monopoly on firei - hur * arms and the soldiery charged with use, the'Neftegna or gunmen. As Caulk (1971) has discussed in detail, Menelik preserved the structure of the administrative and taxation system originated under the amirs and revised and expanded by the Turco-Egyptians. In the region surrounding the city, the control of Qottu-populated agricultural lands was overseen by the damins, malaqs, and garads. Many of these, particularly in the higher offices were Harari. Some were Oromo, the incipient elite whose "emergence" was now at an end. Harari clerks were retained in government offices, where they continued to keep records in Arabic (until 1916, when Amharic was adopted). The basic tax on .prqperty, the.zikat tithe of one-tenth value was renamed asrat and retained, whil e separate tax on productivity was added. Wi'thin the city, Islamic institutions were tolerated, including the recognition of shari'a law for inter-Muslim affairs. Harari land deeds (hojja) provided a legal basis for retaining property, although they were frequently countermanded by the new government's land claims and tenure system. The most immediate social perception of conquest for the Harari, beyond the final extirpation of the ami rate, was the settlement of hundreds of Menelik's troops in the quarter of Harar called Bedroberi, supplanting the families of Harari killed at Ch'elenqo, many or most of whom had come from that quarter. In all , some 12,000 highlanders settled in and around Harar in the years immediately following conquest (Caulk, 1975:6). Land Alienation: Introduction of the Gabbar System The form of land alienation experienced in Harar (including now the Ittu Oromo Ch'ercher massif as well as the region controlled by the Turco-Egyptians) was a regional variant of that discussed in detail by Markakis (1974:104-140). Although it is possible to summar-Tze the pattern this took in the Harar area, no quantitative historical study has yet been published, and therefore its full implication for the political economy of the region TS yet to be revealed. All land was proclaimed crown property, Ras Makonn'en's garrison and officials were provisioned by land confiscated as madbiet ("Kitchen-land"), either under the form of hudad,from which all profits and proceeds went to the treasury, or as maderya,whose income served as the salary of officials (Tsehay Berhane Selassie, . . 1975:1, 7). The princip/<2. of expropriation was that of sisso ("one-third"), wherein the state took two thirds of a parcel of land, and the remainder was left for the original occupants (Markakis, 1975:108). It is hardly surprising that the conquerors typically took the most productive two-thirds. In Harar» a major biqc of land designated as madbiet was the property of the amir which lay beyond the area directly farmed by Harari, and whose cultivators were Qottu. In this and similar confiscations, the peasantry was left on the land, and worked for the crown as its gabbars. Certain lands were all ocated directly as estates for the nobility. A large proportion of the Qottu-farmed lands were entailed in this manner in the Harar region, including the Combo"! cha estate owned by Ras Makonnen and which ' passed intact to his son, Tafari Makonnen, later known as Haile Selassie. At least two rich parcels of land directly outside the city were taken under a third type of alienation, samon grants which supported Christian churches. Throughout the rural regions, more of these church lands were designated as years went by. Markakis illustrates the manner in which the church lands of Asba Tafari became identified with a single highland family, supported by Oromo labor (Markakis. 1974:131-32). fourth form of alienation incurred in Harar was the totality of grants given to the neftegna and the thousands of highland immigrants whose original numbers may have been as high as 12,000 (Caulk, 1975:6) and whose descendants and followers now comprise at least one-half of the population of the administrative district of Harar. Harar Zuria Awraja. The first category of land, madbiet, was originally state possession, not permanently held by individuals. However, much of this was manipulated so as to become yariste gu1t,i.e., inherited land with implied control over the peasant cultivators. Markakis comments on this phenomenon as it progressed throughout Menelik's southern provinces. Through the use of political power, landholding rights were converted from temporary to permanent form; inaccurate measurement enlarged private landhotdings; fraud, intimidation, and force were employed to deprive the southern gabbars^ of their ill-defined rights; and " frequently, outright seizure converted sizeable areas of land into the private property of powerful officials (Markakis, 1974:117). Tsehay Berhane Sellasie has reported this accumulative transfer of property to individuals in the region of Harar. In the following quotation, I believe it is noteworthy that she is referring to usurpations from the most protected of •the indigenous officials, the Harari and Oromo administrators, whose status and land claims had an established and legal basis in the new regime. (Presumably the alienation of lands held by Qottu peasants was accomplished more directly.) The process of transferring those key positions (malaq, garad, etc) to the 'Amhara'...continued without restraint. On various pretences the (officials) began to 'inherit' these positions from the original garadas and malagas. Some acquired the post by paying the bad debts of the-original title holders; others on the promise of taking care of those officers in their old age, and others on the excuse of zemedena or 'friendship' a dubious term suggesting forced acquisition. There was no limit to the number of the garadenat ("garad-ships") or mataqnat ("mataq- ships") a person could ho-ld. in this manner.... Theoretically...and practically this meant that the Amhara began to have only direct administrative power over the land. (Tsehay Berhane Selassie, 1975:4; parentheses added.) Whether the Oromo peasant worked under the indirect rule of the damin hierarchy, or, as was typical farther away from the city. under one form or another of tenancy under the direct control of a member of the new ruling class, the duties of the tribute-payer -- the gabbar—were onerous. First, at, least onethird of the land ' s produce went as tribute to the landholder. Beyond this. and from the remaining production, the gabbar provided the asrat tax, and later a separate land tax. The gabbar was also required to provide manual service, as well as consumption requirements of the landholder, such as meat, honey, firewood, etc. The gabbar was also to r-equrr&d^ -provide a host of other maintenance and exceptional services (Markakis, 1974:114). Muhammed Hassan, specifically referring to the conditions in the Harar region underwrites Markakis' general description of the lot of the gabbar. The peasants in the surrounding towns were given to the Neftagna to be "eaten". The locall y incorporated titleholders also shared in this 'eating' process. As a result of Malkanna (chief of 1,000 men) received from the government from 200-300 Gabbars, a Shambat (chief of 100 men) received 70, a Matoalafa (chief of less than 100 men) 40, a Hsaalaga, 25, and the soldiers, 10 gabbars each. (Muhammed Hassan, 1981:7). As the following statement , by Caulk indicates, conditions became worse as time went on. ...as malkagna were appointed over more and more districts especially after Makonnen's death in 1906, those with less land including the poorer garad were reduced to near serfdom—worse off than slaves, Ras Imru commented in an interview, because they had to work day and night for their maklagna and his wife and then pay taxes besides (Caulk, 1975:13). In the words of the late Oromo poet . Sheikh Mohammad Abdutlahi * this was "the time whan Harar Oromos lived in 'hell' on earth" (Quoted in Muhammed Hassan, 1981:8). - •• In the years immediately following his enthronement, Haile Selassie passed series of reforms which abolished the gabbar's duties in service and kind, as well as protecting their remaining landholdings. Fiscal reforms passed in the 1940*s and later (Markakis, 1974:120), designed to discourage the pocketing of taxes by administrative intermediaries, alsa somewhat lightened the burden of these tenants. In 1968, a government survey of Hararghe Province (National Sample Survey No. 12, 1968) concluded that, "The distribution of holdings by type of land tenure indicates that 46% of the holdings were owned, 49% rented, and 5% partly owned and partly rented from other" (^. 33). -RentiTig implies tenancy, the continuity of the gabbar sys-tem. Hararghe is a vast and ecologically diversified region and the variation in percentage of tenancy is considerable . In the Ch' erch'er region, probably the richest agricultural department of the province , 89% of land holdings were occupied by tenancy, whereas in the district of Harar Zuria Awraja , 89% of the holdings were by owners ( ibid.), many of these Harari . Throughout the Imperial period , local government among the Qottu—the rural Oromo peasantry in the region of Harar—were administered through damins and garads, a variation of the system inherited from the amirate, as Bizuayehu Alemayehu ' s 1972 study of Harar Zuria Awraja documents. These offices , which are administratively the equivalent of the ballabat in other regions, continued to follow the pattern established in the days of the amirate , although the cconomir ^•Ivont .iqos of these positions were much reduced, probably by the fiscal reforms of the 1940 ' s. Thus, in 1972, the damins provided the link between district government and the Orcmo population , and each such damin would,have several garads under his charge. Each village would be represented by a garad. Damins inherited their position , according to primogeniture ( Bizuahyhu Alemayehu, 1972 :55) and received $20 per month . The garads collected taxes in each village and announced government policies . Most spoke some Amharic ( ib^-d.).* The position of Mikitil Worreda Governor (ibid, 54) was the structural equivalent of the damin, and their duties overlapped. However, the Mikitil Worreda Governor was appointed by the Awraja , and was directly responsible to it. Some indication of the overall stratification of ' the region can be deduced from the same work . The author discusses relations between Oromo and government officers, all of whose names are Christian as reported . Significantly , in this study of Harar Zuria Awraja, relations between Oromo and Harari, per se, are not mentioned , and are obviously of secondary significance to Oromo Amhara rela - - tions. This indicates the shift in stratification of the region, except -- as in my early fieldwork --when viewed from the perspective of the Harari minority. Moreover , this report indicates that the gosa ("localised lineage") fulfills the role of a voluntary association,vis - a -vis the government, including organizing and collecting funds for road building ( Bizayehu Alemayehu, 1972: 27). Although the Afran Qallo has long since disappeared , with its central assembly banned by Ras Makonnen and its powers collapsed , the position of Bokku ( representative) has survived, and, as a representative of the gosa is contrasted with the yarad, the local representative of government. Thus, despite the decimation of the Afran Qallo and the enforced impoverishment of the gabbar system, a degree ' of traditional Oromo community organization survived into the 1970' s. Political Economy of Harar i-n the Early 20 th Century By 1900, the political economy of the Harar region had undergone radical changes. The conquest and occupation by Imperial forces, the social consequences of the introduced g' abbar system, had reduced the Qottu peasantry to a stute of servitude. The Harari, whose social system within the city survived the initial stages of occupation, nonetheless were drastically affected by the political and economic ' V changes induced by Imperial rule. As discussed , Harari staffed the -\ clerical positions within municipal offices and served as administrative intermediaries and tax collectors under what was, effectively , a system of indirect rule. In the initial decades of incorporation in the Empire of Ethiopia , their economic fate was ambiguous . Land holdings in certain areas adjacent to the city had been confiscated to provision the garrison, fco reward Ethiopian officials, and to provide Lebensraum for s.oldiers and other highland settlers . Wealthy rn fa xlies, particularly < he amir garacli, had lost most of their outlying farmlands and the income thus derived. On the other hand, the volume of trade passing through the city increased greatly . Although the Turco-Egyptians had secured the routes connecting Harar with the Red Sea, only after Menelik ' s conquest was trade to the interior made dependable. ( After Menelik 's occupation) Harar began to benefit from inte - gration in the Ethiopian state . Robecchi-Bricchetti stated a couple of years later that caravans arrived or left every day in all directions, while Michel at the turn of the century reported that many caravans travelled from Shoa through the city to the coast and that many more Europeans and Armenians had arrived. . . The development of Addis Ababa led to further expansion of trade, as the city was on the capital' s main route to the coast. By 1897 Wylde reported that Harar was ' full to overflowing' and its trade 'rapidly increasing '. . . ( Pankhurst, 1968:410). &" ome Harari merchants undoubtedly thrived during the first two decades of Menelik' s occupation , since trading within Ethiopia at this time was a skill virtually monopolized by Muslims . However, European brokers and tradesmen dealt directly with Menelik 's government in the crucial commodity of firearms, and even they complained about the close control of conditions and terms of this potentially lucrative trade by Menelik. ( Starkie, 1937: 128). Perhaps more important in its long run importance to the Harari than the trade in firearms was the development of a reliable and larqe-volumo inland mar.ket for a host of imported consumer goods: cotton, woolen and silk dry ' joods; sugar, spices, and soap? cooking ware and hardware; perfume and spirits. The total value of such -imports through Harar in 1899-1900 was over 3,800,000 Maria Theresa dollars ( Pankhurst, 1968:415-16). The Indian merchants who flocked to the region handled most of this trade initially. Under the amirate much of the long distance trade of the city had been controlled by amir garach nnd n handful of other wealthy traders. Only with the increased capital facilitated by the augmentation of land holdings under Turco-Egyptian rule was merchandising opened to most Harari, and it is the Indian merchants whom the Harari credit with explaining the rudiments of the merchandising of imported consumer goods . . The Indian merchant Mohammed Alt, who maintained establishments in Harar and Addis Ababa, is . specifically cited as instructing members of the Harari citizenry in these merchandising techniques ( Bab Haji Mure Bashir , Interview, 1964). This initial surge in trade volume, with its consequent effects on the economy of Harar, was permanently ec ' lipsed with the construc - tion of the railroad from Djibouti. Although Harar initially had been on its planned route, the terrain was formidable, and Harar was bypassed . The new town of Dire Dawa became the major customs station and transshipment point, and the new economic hub of the region. The new railroad reached Dire Dawa in December, 1902, and after a chaotic period of financial and diplomatic intrigues, the section connecting Dire Dawa with Addis Ababa wa. s started in 1909 and completed in 1917 ( Pankhurst, 1968;323,334). .Monelik and his successors thus gained the infrastructure for reliable and permanent access to the Red Sea 'V and European imports, but the consequences for Harar were economically disastrous, despite the building of a toll-road connecting it with Dire Dawa ( Pankhurst, 1975:3 ). For the Harari who converted from primary reliance upon absentee landlordism and other farm income to merchandising , the consequences were inevitable and socially disruptive: to pursue their livelihood , ^nore and more of them had to move outside their city to the growing markets of Dire Dawa and Addis Ababa. This exodus gained impetus through the next half century as amajor adaptation of the merchant segment of the Harari population . Changes in the ehtnic make - up of the city of Harar itself indi - cate, in one sense, the significance of conquest and ensuing sociopolitical changes. Paulitschke, who visited Harar in 1886, gave the following figures, based on a Turco-Egyptian census . In 1885 the total population of Harar was 42,000, including, the 3411 men of the Egyptian garrison; 24-25,000 "pure Harari", some 6,000 Oromo, 5,000 Somali, and - most interestingly - ". . '. 3,000 almost wholly denationalized ("entnationalisirte ") Abyssinians from Shoa and Amharas" ( Paulitschke , 1888:207-208; my translation }.. The ambiguity of the description of the highlanders here is intruiging . ( Did this three thousand include Muslim traders, who might have been Argobba, or did it describe only highland Christians?) survey of language affinity in the city of Harar ( including the new area outside the city wall, a region* at least as large as old city), states that 57.2';, of the population of 41,150 in 1965 spoke Amharic as its mother - tongue, 8.6 'c spoke the Oromo language, and 31.1% spoke "other Ethiopian languages", most of which was presumably Harari. Numerically , only 12,820 persons fell into this •*• category ( Ethiopian Government. Cen-fcral Statistical Office, 1967 : 6). Three inter-related conclusions are easily drawn from the two sets of data. The presence of Amharic speakers became numerically dominant (as well as politically and economically dominant) during the first three-quarters of a century within the Empire. - The total city population remained remarkably stable, although one must realize that the "new" part of the city, much of which-was built during the 1936-41 Italian rule, occupies an area as large as the old walled city . A somewhat startling fact is that the Harari population dropped by fifty percent during this period . Revolution & Reform: Impact on Harari & Oromo Following the deposal of Haile Sellassie in 197*/, the reforms of t.he revolutionary government of Ethiopia, the PMAC or Derg, . irreversibly altered the political economy of Ethiopia, IIarar Province included. A much - needed land - reform was effected in a single sweeping act, the "Proclamation to Provide for the Public-Ownership of Rural Lands", April, 1975. Further levelling of disparities in the distribution of wealth in Ethiopia was effected by the "Government Ownership of Urban Lands - and Extra Houses Proclama - tion," July 1975. As essential as these reforms undeniably were for eradicating the heritage of the gabbar system and other exploitative tenancy in Ethiopia, these reforms spelled disaster for the Harari who remained in the old city. The land reforms, which limited family ownership to ten hectares, eliminated the extensive landholdinqs that- many Harari had accumulated and protected through the ami rate, the Turco-Egyptians, and the Empire. This wealth in land provided the oldesL and strongest pillar of the old city' s economy, and thus supported its highly spe- -y cialized way of life behind the city wall Harari community efforts - - to adapt to this drastic transition are discussed elsewhere ( Maldron, m b S). ^ ^ For the surrounding Oromo, the immediate reaction to the revolutionary government's land reform was relief, if not elation . Oromo labor , some of it unpaid, had supported the urban elite of Harar since the beginnings of sedentization and peasantisation in the region. Nineteenth-century patterns of labor exploitation in the hands of the Harari and Oromo elite had been expanded by the Turco Egyptians and doubled in severity by the "Amhara^. The Land Reform promised to return Oromo lands to the laborers, but the announcement of the reform was counterbalanced by a new presence of government with further plans for organizing and utilizing Oromo labor - this time in peasant associations - to support the State eccx^Otr^y* The impact of the land cmd property reforms on the relationship between Harari and Oromo was of secondary importance to its effect on each group ' in relationship to the national government - However, there is no doubt that with the elimination of absentee landlordism in the region, the privilege of Harari versus Oromo- was also ended. The remnant stratification which had served as a rationale for mediating Uo Harari boundary-management practices with the Oromo was^ ' *ended, but more- important than this vestige of the amirate was the impact of the reforms on the city itself. In Harari eyes, the urban property reform had eliminated their city sanctuary. » The Harari at this fcimr were in a subtly strange position. They were actively resented }'y many Qottu Oromo and other local Oromo as exploiters. Some Harari served the malkagna system under the Empire as indirect rulers, under the damin system, and in many Qottu areas .. . they were middlemen who became the focus of resentment s However, Harari citizens had themselves emerged from an understratum background, albeit a comparatively comfortable- urban one. They had succeeded in adjusting to tlie rule of outsiders by providing essential clerical services, and they had gained some property by using these positions astutely. Yet they had shared with the Oromo the loss of political sovereignty and control of their political fate. Whereas the Oromo were disenfrachised by patterns of tenancy under the Empire, the Harari disenft-achisement came about as the result of much needed reforms in Ethiopia. In 1975, having lost their farms and urban properties, the Harari in the old city viewed themselves as victims. Conclusions: Harari-Oromo Ethnic Stratification Reconsidered 1. A simple Harari-Oromo ethnic stratification , in terms of one ethnic-class ' s domination of another such group' s life by virtue of its control of basic resources, never truly existed in and around Harar. At the end of the emirate, an "incipient stratification" existed ( see p. l^ }. Centralization of power occurred only with the appearance of the Turco - Egyptians, who con - trolled the Oromo. The Harari citizenry occupied a privileged position in this ten-year period , but they did not control the system. /, Harar as thecanter of 2. With Ethiopian occupation,^a. political - economic system, rationalized in terms of inter -class/inter-»ethnic status differentials , ceases to exist. The focus of the subject dissolve's and merges into the greater frame of Ethiopian expansion and ytate formation. Harari-Oromo relations iir-o affected, of course, but they are completely secondary to the structural position assumed in the Haraa/ region by the intrusive Ethiopian government and personnel . 3. What, finally, is the evaluation of the model of ethnic stratification as earlier presented? ( a) The drastic alteration of the local political economy with the incorporation of Harar into the Empire meant that only a "remnant stratification" remained. That is, the Harari were able to retain a restricted position of absentee landlordism , utilizing Oromo labor, but always losing ground. ( b) By the 1960 *s, the model of ethnic stratification was certainly not based on political-economic reality , but was an ideological device of the Harari. ( i) By retaining an idealized concept of their Simplified superior position versus the Oromo , they were able to provide a set of boundary -management dicta which , by referring to traditional, symbolic "outsiders", enabled ^ them, as an encapsulated minority population under increasing pressures of dissolution , to maximize forces of cultural conservatism which fortified their cultural integrity , ( ii) The greatest pressures upon Harar society , in the 1960 's, resulted from economic, demographic, social, and political pressures of change directly or indirectly associated with the incorporation into Ethiopia and Ethiopia' s modernization. The Harari model of ethnic stratification, as presented in the 1960 ' s, then, was an ideological attempt to cope with these pressures by using an archaic pattern of idealized boundary - maintaining prescrip - tions. To put thir mor< clearly, the Harari were attempting to utilize an old modc'l of iIarari-Oromo ethnic stratification symbolically, as a tool for coping with pressures of change accruing from all that being part of 20 th century Ethiopia 'y entailed . 4. Ethnicity as epiph.enomenon. Anthropologists are sometimes assailed for concentrating on ethnic groups and ethnicity, and by doing so, fostering policies of "divide and rule". By now in this discussion some implied directions for future analysis of Harari-Oromo ethnicity and ethnic relations are apparent. ( a) Ethnic relations are probably best viewed as long-term outcomes of contact and exchange situations, where groups differentiate in separate but interdependent ecological or economic niches. The perception of intergroup differences and status differentials are likely to be rationalized in terms of ethnic stereotypes, particularly where, as in this case, each group has its own language and customary institutions , (b) Nonetheless, the major conclusion is that ethnic relations change in response to underlying political and economic determinants which change the relationship of the contact groups, or alter the control of a group upon its niche. -Ethnic relations are, in sum, after - the - fact rationalisations more than they are Ding -an-sich. Ethnic relations can be changed by changing the underlying principles of groups ' differentiation, and these are distinctly po - litical-economic. 5. Finally, from an analytfc standpoint, the major stratification of the Harar region after its incorporation into the Empire is but a local version of the overall pattern experienced by the southern provinces .lurin i the century of Imperial rule. This is a true stratification, where access to power was monopolized by the Imperial government and its personnel and the labor of the system was provided by suppressed Oromo populations . The local variation of Harar Province was that the Harari served in '• ballabat-like capacity in and around the old city, and in the ' * r process were reduced to a virtually powerless middlemen minority , somewhat susceptible to scapegoating . The unanswered , but most important, question of the entire discussion involves not the interesting but minor problem of Harari-Oromo relations, but the imposed stratification in the southern provinces between Empire and Oromo. If , I suggest, the Promo perception of this exploitative relationship is in terms of an historical government dominance, per se, and an unfair power differential between ruled and ruler, then the increasingly self-aware bloc of Oromo, within Ethiopia may find adequate satisfaction and representation within a revolutionary government , wherein their participation is a class phenomenon . However , if the Oromo perception of the true stratification which developed under the Empire in the Southern Provinces, including Harar, is perceived as primarily an ethnic stratification, where their oppression is blamed on a generalized "Amhara" occupier, rather than the Imperial government, then a separatist •Oromo nationalism is more likely to emerge. The present study may aid those involved in solving such problems by suggesting that such movements are motivated by political and economic imbalances, and can be adjusted temperateiy by addressing such incqucili ties, rather Llinn by attacking ethnicity itself. • « REFERENCES Abir, Mordechai 1980 Ethiopia and the Red Sea . London ( Frank Cass) Bloch, M. 1963 "The social influence of salt." Scientific American 209 ( 1): 88-89. Braukamper, Ulrich 1979 "Islamic principalities in southeast Ethiopia between the 13 th & 16 th centuries," Ethiopianist Notes 2 ( I): 17-56. Burton, Richard 1966 ( orig. 1856) First Footsteps in East Africa London ( Routledge & Kegan Paul) . Caulk, Richard A. 1971 "The occupation of Harar: January 1877," Journal of Ethiopian Studies IX (2): 1-20. 1975 "Menelik' s conquest and local leaders in Harar," unpublished paper. Conference, History & Culture of the Peoples of Harar Province. Institute of Ethiopian Studies, Addis Ababa . 1977 "Harar town and its neighbors in the nineteenth century," Journal of African History XVII ( 3): 369-^6. Cerulli, Enrico 1936 Studi Etiopici I: La Lingua e La Storia di Harar Rome (Istituto per L'Oriente). 1942 "Document! arabi per la storia dell' Etiopia," Rendiconti delta Reale Accademia del Lincei, ser iv, IV: 2 20. - Cohen, Abner 1974 Urban Ethnicity London TTaviTStock Publications) . 4 Ethiopia, Central Statistics Office 1967 Report on a Survey of Harar. Addis Ababa (C. S.O.) 1968 Report on a Survey of Hararge Province. National Sample Survey No. 12. Addis Ababa ( -C.S. O.) Ejetta Feyassa Basha 1975 "Newcomers andrthe peoples of Harar in the early 20 th centur. y," unpublished paper. Conference, History & Culture of the Peoples of Harar Province. - E-n' stitute oi: Ethiopian Studies, Addis Ababa Fried, Morton 1957 The Evolution of Political Society. New York ( Random House) Hetzron, Robert 1972 Ethiopian Semitic: Studies in Classification Journal of Semitic Studies Monograph no. 2. Cambridge ( Manchester University Press) Koehn, Peter & Sidney R. Waldron 1978 Afocha: A Link Between Communi.ty and Administration in Harar, Ethiopia Maxwell School, Foreign & Comparative Studies/Africa XXXI Syracuse ( NY) ( Syracuse University Press). Legesse, Asmarom 1973. Gadai Three Approaches to the Study py an African Society Mew York (Free Press). Marcus, Harold G. 1975 The Life and Times of Manelik II: Ethiopia 1844 - 1913 Oxford (Clarendon Press) Markakis , John 1974 Ethiopia; Anatomy of a Traditional Polity Oxford ( Clarendon Press). ?Muhammad Hassan 1973 "The relation between Harar and the surrounding Oromo between 1800-1887." Unpublished thesis. Department of History, Haile Selassie I University, Add' is Ababa. /* 1 Q p j "Menelik ' s conquest of Harar, 1887, and its effect on the political organization of the surrounding Oromos up to 1900," unpublished paper. Conference on the History & Society in the South Ethiopian Periphery, Stanford University, March , 1981. Pankhurst, Richard 1968 Economic History of Ethiopia, 1800-1935 Addis Ababa ( Haile Sellassie University Press). 1975 "The Harar - Dire Dawa road and the ' Road Construction Association * of 1934. unpublished paper. Conference, History & Culture of the Peoples of Harar Province. Institute of Ethiopian Studies, Addis Ababa. Paulitschke, Phillipp 1888 Harar: Forschungsreise nach den Somal - und Galla - landern Oat AJJiikaa .- Leipzig ( F. A. Brockhaus) . Sahlins, Marshall 1972 Stone Age Economics Chicago ( Aldine) . Shihab ad-Din, Ahmed ( Arab Fuqih) 1897 Futuh al-Habasha: Histoire de la Conqu^te de l"Abyssinie ( trans. & ed. R. Basset) Paris Sjoberg, Gideon 1960 The Preindustrial City Glencoe ( Free Press). Starkie, Enid 1937 Arthur Rimbaud in Abyssinia Oxford- (Clarendon Press) Stitz, Volker 1975 "Arabic town records and the economic and population history of Harar during the 19 th century," unpublished paper. Conference, History & Culture of the Peoples of Harar ProviJice . Institute of Ethiopian Studies. Addis Ababa . Trimingham, J. Spencer 1965 Islam in Ethiopia , 2 nd ed. London & Liverpool ( Frank Cass). Tsehay Berhane Selassie 1975 "Development of land tenure in Hararge Province, 1887-1975," unpublished paper. Conference, History & Culture of the Peoples of Harar Province. Insitute of Ethiopian Studies. Addis Ababa. Waldron, Sidney R. 1974 "Social organization and social control in the walled Muslim city of Harar, Ethiopia," unpublished PhD thesis, Department of Anthropology , Columbia University . 1979a "Harar: the Muslim city in Ethiopia," Proceedings of the Fifth International Conference on Ethiopian Studies, , (ed,R. Hess) Chicago ( University of Illinois). 1979b "A farewell to Bab Haji: City symbolism and Harari identity, 1887-1977," unpublished paper. Workshop,- Society & History in the South Ethiopian Periphery,African Studies Centre, University of Cambridge. 1980 "Within the wall and beyond: Harari ethnic identity and its future," in Urban Life: Readings in Urban Anthropology < ed. G. Gmelch & W. Zenner). New York ( St Martin 's Press) Yusuf Ahmed 1961 "An inquiry into some aspects of the economy of Harar and the records of the household economy of the amirs of Harar ( 1825-1875)" University College of Addis Ababa, Ethnological, Society, Bulletin ( 5) Yusuf Ahmed Wuber 1965 "Afocha" Journal of Ethiopian Studies4:125-32.